Amnon Sonnenberg1. 1. The Portland VA Medical Center and the Division of Gastroenterology/Hepatology, Oregon Health Sciences University, Portland, Oregon 97239, USA.
Abstract
OBJECTIVES: Although the tools of medical decision analysis have been mostly used to address complex issues of healthcare and management policies, an important role remains for decision analysis to play in solving mundane medical problems of daily practice. The present article aims to restate the origins of decision analysis and to demonstrate the utility of decision models in clinical gastroenterology. METHODS: Three clinical scenarios are presented to illustrate the applicability of decision analysis in clinical gastroenterology. The examples are modeled as simple decision trees and phrased in terms of threshold analysis. RESULTS: In spite of its striking simplicity, the framework of threshold analysis is able to capture a large variety of heterogeneous and often perplexing medical problems. Rather than calculate a cost-effectiveness or cost-benefit ratio, the result of the analysis is expressed as a probability value, which forms the threshold between the two choices against or in favor of a particular medical action. For most analyses little if any calculation is needed to solve the decision tree, and the final outcome of the analysis can be derived based on comparative estimates and considerations of magnitude only. CONCLUSIONS: The study demonstrates that threshold analysis can serve as a simple and convenient instrument to solve a large variety of problems in clinical gastroenterology.
OBJECTIVES: Although the tools of medical decision analysis have been mostly used to address complex issues of healthcare and management policies, an important role remains for decision analysis to play in solving mundane medical problems of daily practice. The present article aims to restate the origins of decision analysis and to demonstrate the utility of decision models in clinical gastroenterology. METHODS: Three clinical scenarios are presented to illustrate the applicability of decision analysis in clinical gastroenterology. The examples are modeled as simple decision trees and phrased in terms of threshold analysis. RESULTS: In spite of its striking simplicity, the framework of threshold analysis is able to capture a large variety of heterogeneous and often perplexing medical problems. Rather than calculate a cost-effectiveness or cost-benefit ratio, the result of the analysis is expressed as a probability value, which forms the threshold between the two choices against or in favor of a particular medical action. For most analyses little if any calculation is needed to solve the decision tree, and the final outcome of the analysis can be derived based on comparative estimates and considerations of magnitude only. CONCLUSIONS: The study demonstrates that threshold analysis can serve as a simple and convenient instrument to solve a large variety of problems in clinical gastroenterology.